January 1998 - techniques

• Camera Techniques- The overlooked Tools! - Flat Plate CameraBruce Fox and 2 dimensional plate techniques Kath Clay
• Sample Holders Tim Hyde
• Reitveld 'R' values discussed

Cover Photograph - Flat Plate Camera in Action

Camera Techniques - The overlooked Tools!

A flat plate photograph showing a wealth of information obtained quickly and simply.

�The specimen is a thin plastic sample consisting of two main inorganic fillers in a polymer matrix, together with a variety of other minor additives. The two fillers each have two polymorphs and the polymer is semi-crystalline, resulting in five crystalline phases present. There are, inevitably, overlapping peaks, and one of the inorganic polymorphs has a needle-like habit. As a result of processing, the polymer crystallites orient on three axes and the needle like inorganic polymorph conforms to these orientations. A single flat plate photograph allows for the phase identification of the four polymorphs with a surprising degree of accuracy; a description of the fibre orientation of the polymer and needle-like polymorph; and a qualitative estimate of the crystallite sizes of the various phases. Time: 1 hour exposure, 10 minutes development, 30minutes measurement and data reduction. This is just one of several invaluable camera techniques we have set up, which are ideal for screening quite complex materials in a simple and rapid way.

Bruce Fox

Two dimensional plate methods, should they be more widely used?

Diffractometer based, computer controlled, digital collection of polycrystalline diffraction data is accepted as the industrial norm, a rapid and cost effective analysis tool. In fact, many younger practitioners in industry have never even seen, let alone analysed plate, Debye Scherrer or equivalent films. Frustratingly, automated phase identification routines of diffractometer based diffraction traces sometimes produce what the analyst considers to be 'spurious' identifications, when considering other information available (e.g. origin, chemistry, physical properties). These conundrums generally occur when the specimen is of very fine particle size, has coarse grains, is strained or has preferred orientation, (as either the diffraction peak's shape, position or intensity are modulated).

�Whilst all these phenomena can be investigated by specialist techniques, requirin g dedicated software and / or hardware, the overview approach of looking at the characteristics of the Debye ring is frequently ignored. Why? Because plate film or film cameras are not now a standard set up in mainly industrial laboratories, (they are not High Tech!) Yet, as shown above, films can provide a rapid structural information overview and pointers to the selection of the most appropriate specialist technique for a particular problem. Further, 'spurious' automated phase identifications can easily be proved and explained. A summary of the information available from the intensity features of a Debye ring is provided below.

  Appearance of Diffraction Line  Condition of Specimen

       Continuous                  fine grained 

                                     (or coarse grained and cold-worked)

       Spotty                      coarse grain

       Narrow                      strain free

       Broad                       residual strain 

                         

            and/or small particle size

       Uniform Intensity           random orientation

       Non-uniform Intensity       preferred orientation

So, for the little extra effort required to set up a plate film or film camera facility in your laboratory, you may be able to enhance the cost effectiveness of your resource!

Kath Clay

Help - Sample Holders

When did you last buy new s ample holders? This question is particularly relevant to Bruker AXS (formerly Siemens) D500 kit with an automatic sample changer unit, for which I have first hand experience, but may also be of relevance for your diffraction equipment too.

As with most of the diffraction community we religiously measure instrumental standards at least weekly to check sample alignment errors, tube intensity etc. These results were telling us that there was no long term change in anything except the expected decline in tube intensity.

Results on experimental samples however were suggesting otherwise. At first we thought peak shifting from literature positions was perhaps a feature or function of our synthesised high surface area catalysts or from added or adventitious dopants. However when pure high temperature calcined (experimental) standards gave unusually large lattice parameters we needed a more practical explanation.

Of course with hindsight its easy to solve. Throw your old sa mple holders away and buy some more. Years of use and subsequent cleaning had lead to a wearing away of minute amounts of the plastic material of the sample holder totally destroying sample alignment. And worryingly at the time we had no way to monitor this as all our instrumental standards are mounted (fixed) in sample holders that never received the same degradation from day to day handling treatments as our routine sample holders.

Tim Hyde

The 3 R's - Rietveld R values discussed

The Rietveld method has become a well-established technique of analysis for powder diffraction data. Least squares minimisation (refinement) of the difference between observed data and that calculated from a model structure is used to produce estimates of variables such as lattice parameters, preferred orientation and atomic co-ordinates. Assessment of the fit between the calculated and observed data is most often throug h calculation of ‘R-values’. The precise meaning of these R-values and their use and abuse has resulted in some recent debate.

�Perhaps the 3 most frequently used R-values are :

R_B :

R-Bragg is produced from the Bragg intensities calculated from the structural model and those estimated from the observed data. Hence, this R-value does not truly reflect the fit to the actual observed data points but is closely related to the conventi onal R-values from single crystal refinements where a ‘good’ refinement produces R_B's of < 5%.

R_wp :

R-weighted pattern is calculated from the differences between each individual point of the observed and calculated powder diffraction patterns. (If weights are 1.0 then R_p is calculated). It is perhaps the most meaningful R-value and is a good indicator of the progress of the refinements. However, even when the structural model is good, small anomalies in the observed data (e.g. peaks from a trace second phase) can significantly inflate R_wp values. Further, small values of R_wp can be produced from a poor structural model by fitting low resolution data or data that is not sensitive to structural parameters. Values of R_wp reported for refined structures are often in the range 5 - 15 %.

R_e:

R-expected may be thought of as a ‘theoretical' minimum R-value that can be achieved with the observed data and structural model variables. Thus, unlike the other R-values, its numerical value does not change during the course of a single refinement. The ratio, R_wp/R_e is frequently used as a ‘goodness-of-fit’ indicator (sometimes c), which reflects an adequate refinement when its value is 1.0 to 1.3. For values >1.3 the structural model should be reconsidered and for values <1.0 the quality of the data sho uld be examined.

�Recent debate has concerned the effect that the background intensity distribution has upon the Rietveld refinements. Most workers refine some background function as part of the minimisation process. However, Hugo Rietveld’s original work was based upon observed data which had been background subtracted. Thus there is a distinction between conventional R-values (without background) and the more frequently quoted, total R-values (including backg round).

�Whichever R-values are most appropriate, there can be no substitute for an intelligent inspection of the calculated and observed patterns, and close examination of the final structural model parameters (and their errors) to ensure that they are sensible.

For more detailed description and discussion the reader is directed to :

1. The Rietveld Method ed. R.A. Young IUC, OUP 1993
2. The Internet Frequently Asked Questions

Keith D Rogers